How to make my own classes with C++ and SFML

Question

I have currently started working with SFML after learning the basics of C++. I have learnt about Arrays, References and everything that comes before it but have struggled to grasp the concept of using classes.

In SFML I have created a simple sprite movement program but, I would like to move this information into a class (lets say it will be called "Player"). I have messed around a lot but I can not get it to work.

I have tried creating a function in a class that would check for player input, but I can not access my sprite that I created in main. I would like to move everything related to the player into a Player class but need some advice.

What is the correct way to do this? (Please don't say go back and learn about classes, this is where I want to learn about them!)

main.cpp

#include <SFML/Graphics.hpp>
#include <string>
#include <iostream>

int main()
{
    //character position
    enum Direction{ Down, Left, Right, Up };
    sf::Vector2i source(1, Down);

    //window
    sf::RenderWindow window(sf::VideoMode(1200, 700), "Testing");
    window.setKeyRepeatEnabled(false);

    //player character
    sf::Texture pTexture;
    sf::Sprite pSprite;
    if(!pTexture.loadFromFile("image/playerSprite.png"))
        std::cout << "Texture Error" << std::endl;
    pSprite.setTexture(pTexture);
    pSprite.setScale(1.5f, 1.5f);

    //game loop
    while (window.isOpen())
    {
        sf::Event event;
        while (window.pollEvent(event))
        {
            if (event.type == sf::Event::Closed)
                window.close();
        }

        window.clear();

        if (sf::Keyboard::isKeyPressed(sf::Keyboard::Up)) //move up
        {

            source.y = Up;
            pSprite.move(0, -0.2);

            //animation
            source.x++;
            if(source.x * 32 >= pTexture.getSize().x)
            {
                source.x = 0;
            }
        }
        else if(sf::Keyboard::isKeyPressed(sf::Keyboard::Down)) //move down
        {
            source.y = Down;
            pSprite.move(0, 0.2);

            //animation
            source.x++;
            if(source.x * 32 >= pTexture.getSize().x)
            {
                source.x = 0;
            }
        }
        else if(sf::Keyboard::isKeyPressed(sf::Keyboard::Right)) //move right
        {
            source.y = Right;
            pSprite.move(0.2, 0);

            //animation
            source.x++;
            if(source.x * 32 >= pTexture.getSize().x)
            {
                source.x = 0;
            }
        }
        else if(sf::Keyboard::isKeyPressed(sf::Keyboard::Left)) //move left
        {
            source.y = Left;
            pSprite.move(-0.2, 0);

            //animation
            source.x++;
            if(source.x * 32 >= pTexture.getSize().x)
            {
                source.x = 0;
            }
        }

        pSprite.setTextureRect(sf::IntRect(source.x * 32, source.y * 32, 32, 32));
        window.draw(pSprite);
        window.display();

    }

    return 0;
}

Answer 1

Disclaimer: You shouldn't expect that kind of answer, you really should read more on OOP to get the point, this has nothing to do with SFML, this is just basic refactoring.

How to think with OOP

First thing first, before coding a feature, you should design the OOP structure that really suits the situation. See each class as part of a whole, that is your program. A class in fact is just an aggregation of data with useful methods that only affects the data inside the class (or the data provided via method parameters) in a meaningful way.

See the basics of C++ (more the OOP part for you) to understand how to get it to work in C++. The concepts are similar in other programming languages.

Working with your provided code

What you asked for was a Player class and it's a great idea to get the player code out of the program main logic. You need to ask yourself: "What my player code needs to work?"

The player class

Basically, your player is only a sprite and a position. So you encapsulate those data into your Player class as private members. That keeps other code from messing with the player data. To use the player data, you need to provide methods in the class that each affects only the Player.

Texture and Sprite

I have kept the Texture outside of the player on purpose. Textures are heavy objects, that's why the Sprite object only keeps a pointer to it. Sprites are lightweight and can be changed and copied easily. The managing of texture objects and other assets is another subject, though here's my own resource manager code.

Optional

I did not took the time to change your code much, but you could change the way you handle the movement to only make one "move" method that takes a Player::Direction has a parameter.

To help you a little more and to give you some more guidelines on the subject, I used "forward declaration" and moved your Direction enum inside the class. It's maybe not the best way to achieve what you want, but I've only change your own code to avoid getting you lost.

The Code

Anyway, here's my go at this.

Player.h

#ifndef PLAYER_H_
#define PLAYER_H_

#include <SFML/Graphics/Drawable.hpp>
#include <SFML/Graphics/Sprite.hpp>

// Forward Declaration
namespace sf {
class Texture;
}

// provide your namespace to avoid collision/ambiguities
namespace test {
/*
 *
 */
class Player: public sf::Drawable {
public:

    enum Direction {
        Down, Left, Right, Up
    };

    Player(const sf::Texture& playerTexture);
    virtual ~Player();

    virtual void draw(sf::RenderTarget& target, sf::RenderStates states) const;

    void moveUp();
    void moveDown();
    void moveLeft();
    void moveRight();

private:

    sf::Sprite mSprite;
    sf::Vector2i mSource;

};

} /* end namespace test */
#endif /* PLAYER_H_ */

Player.cpp

#include "Player.h"

// you need this because of forward declaration
#include <SFML/Graphics/Texture.hpp>
#include <SFML/Graphics/Rect.hpp>
#include <SFML/Graphics/RenderTarget.hpp>

namespace test {
Player::Player(const sf::Texture& imagePath) :
                mSprite(imagePath),
                mSource(1, Player::Down) {

    // do not need that line anymore, thanks to initialiser list
    //pSprite.setTexture(pTexture);

    mSprite.setScale(1.5f, 1.5f);

}

Player::~Player() {
    // TODO Auto-generated destructor stub
}

void Player::draw(sf::RenderTarget& target, sf::RenderStates states) const {
    target.draw(mSprite, states);
}

void Player::moveUp() {
    mSource.y = Up;
    mSprite.move(0, -0.2);

    //animation
    mSource.x++;
    if (mSource.x * 32 >= (int) mSprite.getTexture()->getSize().x) {
        mSource.x = 0;
    }

    mSprite.setTextureRect(sf::IntRect(mSource.x * 32, mSource.y * 32, 32, 32));
}

void Player::moveDown() {
    mSource.y = Down;
    mSprite.move(0, 0.2);

    //animation
    mSource.x++;
    if (mSource.x * 32 >= (int) mSprite.getTexture()->getSize().x) {
        mSource.x = 0;
    }
}

void Player::moveLeft() {
    mSource.y = Left;
    mSprite.move(-0.2, 0);

    //animation
    mSource.x++;
    if (mSource.x * 32 >= (int) mSprite.getTexture()->getSize().x) {
        mSource.x = 0;
    }
}

void Player::moveRight() {
    mSource.y = Right;
    mSprite.move(0.2, 0);

    //animation
    mSource.x++;
    if (mSource.x * 32 >= (int) mSprite.getTexture()->getSize().x) {
        mSource.x = 0;
    }
}

} /* end namespace test */

main.cpp

#include <SFML/Graphics.hpp>
//#include <string> // not used for now
#include <iostream>

// don't forget to include your own header
#include "Player.h"

int main() {

    // just to save typing the "std::"
    using std::cout;
    using std::endl;
    using std::cerr;

    //window
    sf::RenderWindow window(sf::VideoMode(1200, 700), "Testing");
    window.setKeyRepeatEnabled(false);

    //player texture
    sf::Texture pTexture;
    if (!pTexture.loadFromFile("image/playerSprite.png")) {
        cerr << "Texture Error" << endl;
    }

    test::Player thePlayer(pTexture);

    //game loop
    while (window.isOpen()) {
        sf::Event event;
        while (window.pollEvent(event)) {
            if (event.type == sf::Event::Closed) {
                window.close();
            }
        }

        window.clear();

        if (sf::Keyboard::isKeyPressed(sf::Keyboard::Up)) //move up
                {
            thePlayer.moveUp();
        } else if (sf::Keyboard::isKeyPressed(sf::Keyboard::Down)) //move down
                {
            thePlayer.moveDown();
        } else if (sf::Keyboard::isKeyPressed(sf::Keyboard::Right)) //move right
                {
            thePlayer.moveRight();
        } else if (sf::Keyboard::isKeyPressed(sf::Keyboard::Left)) //move left
                {
            thePlayer.moveLeft();
        }

        window.draw(thePlayer);
        window.display();

    }

    return 0;
}

Other good practices

Accessors, or Getters/Setters, are member functions that gives one the access to a class private member.

In your code, you could do something like that:

 class Player {
    public: 
        Player(const sf::Texture& playerTexture);
        virtual ~Player();

       // to give access to a const reference of the sprite
       // One could call it like: sf::Sprite mySprite = myPlayerObject.getSprite();
       // notice also that the method itself is const, which assure you that
       // myPlayerObject won't change by calling getSprite()
       const sf::Sprite& getSprite() const{
           return mSprite;
       }

       // setSprite is not a const method, so it will change the data
       // inside myPlayerObject
       void setSprite(const sf::Sprite& newSprite){
           mSprite = newSprite;
       }

    private:
        sf::Sprite mSprite;
        sf::Vector2i mSource;
    };